PSI - Issue 19

Kai Schnabel et al. / Procedia Structural Integrity 19 (2019) 442–451 Author name / Structural Integrity Procedia 00 (2019) 000 – 000

444

3

2. Investigation of the additively manufactured geometries Flat specimens with a thickness of = 2 and notches with two different stress concentration factors of = 1.69 and = 2.54 where designed and manufactured with a ConceptLaser M2 system, Figure 1. A standard parameter set with different core and contour parameters was used, which is typically used for industrial applications. As base material, AlSi10Mg powder was chosen. The metallographic investigation shows a significant difference between the nominal CAD and the additively manufactured specimen geometry, especially in the notch with an approximately factor of three in the size of the notch radius. Although both specimen were manufactured with the same parameter set, a completely different distribution of pores is present in the area of the notches. The accumulation of pores on one side of the flank of the notch can be traced back to the so-called downskin angle, for which different local process parameters are used Nominal geometry Real notch geometry Nominal K t = 1.7 Nominal K t = 2.5 3. Finite element models As seen in the metallographic investigation, a high deviation between the ‘as - built’ and the nominal geometry results from the use of standard parameter sets. In order to investigate the influence of these differences in terms of cyclic strength and as a basis for further fatigue assessments, three-dimensional idealized FE models without modelled pores are set up taking advantage of geometrical symmetries, Figure 2. The axial load is = 100 . The linear-elastic material parameters are set as homogenous with a Y OUNG ’ S modulus of = 72 and a P OISSON ' S ratio of = 0.33 . Hexaeder 20-node quadratic brick elements with quadratic shape function are used. The notch radius as well as the notch opening angle are varied in order to cover the nominal and the ‘as - built’ geometries. In addition, a parameter study was performed to investigate the influence of the notch radius and the notch opening angle on the stress gradient, Figure 3. It shows that the stress gradient = ∆ 1 ∙ (1 − 2 1 , , ) as well as the stress concentration factor K t are very sensitive to radius changes. A slight decrease of the radius from = 0.3 to = 0.2 increases the stress concentration factor to Δ = 18% , Figure 3 left. Also the stress gradient is increasing. Figure 1 Nominal geometries of the investigated specimens and cross-section of the real geometries with a nominal thickness t = 2 mm